GPT 6.1 Sol API on Sprelay
Use model ID gpt-6.1-sol with an OpenAI-compatible client. Check the live pricing page, then create a dedicated key in a compatible group.
Connection settings
| Setting | Value |
|---|---|
| Model ID | gpt-6.1-sol |
| Client base URL | https://sprelaytoken.com/v1 |
| Chat endpoint | https://sprelaytoken.com/v1/chat/completions |
| Group listed in the public catalog | ChatGPT High-Speed |
These settings were checked in Sprelay's public catalog on 11 October 2026. A catalog listing does not guarantee current upstream availability or compatibility with every client. This guide does not claim a completed authenticated inference test.
Pricing and a small cost example
The live page showed these dated example rates for ChatGPT High-Speed, in USD per 1 million tokens:
| Input | Output | Cached input |
|---|---|---|
| $0.26 | $1.30 | $0.013 |
For 10,000 uncached input tokens and 2,000 output tokens, the illustrated cost is 0.01 × $0.26 + 0.002 × $1.30 = $0.0052.
This excludes extra agent requests and other billable usage. Rates can change; the current price page and actual usage record take precedence. Read billing before scaling.
Make and verify the first request
- Create a dedicated key, set a small quota and choose ChatGPT High-Speed if available to your account.
- Follow the minimal Chat Completions request, replacing its model value with
gpt-6.1-sol. - Send a short prompt. Confirm a successful HTTP response with generated content in
choices. - Check usage logs for the expected key, model, status and charge.
For desktop chat, use the Cherry Studio guide. For Cursor, follow the custom base URL guide and check whether your version supports the intended feature and model.
Troubleshooting
For 401, check the complete key, expiry and balance. For a model access error, verify the group and exact ID. For 404, check that the client has not duplicated /v1. See API errors.
If contacting Sprelay support, provide the timestamp, model ID, status and redacted request ID. Never send your full key.